基于轴距预瞄的半主动悬架模糊神经网络控制

Fuzzy Neural Network Control of Semi-active Suspension Based on Wheelbase Preview

应用模糊控制和神经网络控制理论,构建了1/2车辆的半主动悬架模型,设计了基于轴距预瞄的半主动悬架模糊神经网络控制系统。对前轮半主动悬架采用以对应处车身垂向加速度为目标的模糊控制,对后轮半主动悬架采用轴距预瞄模糊控制,并利用神经网络来调整模糊控制器的控制规则和隶属度函数。在不同车速下对所建的控制系统分别进行了白噪声和路面脉冲输入的仿真。结果表明,与传统的被动系统相比,轴距预瞄模糊神经网络控制的半主动悬架系统能有效降低车辆振动;与模糊控制的半主动悬架系统相比,质心垂向加速度和后轮对应处车身加速度均有显著减小,较好地改善了车辆的行驶平顺性。

By applying theories of fuzzy control and neural network control,a semi-active suspension model for half vehicle is constructed and a fuzzy neural network control system for semi-active suspension is designed based on wheelbase preview.Fuzzy control is applied to front wheel semi-active suspension with the vertical acceleration at the point on car body vertically corresponding to front wheel center as control objective,while the fuzzy control based on wheelbase preview is applied to rear semi-active suspension,and neural network is adopted to adjust the control rules and membership function of fuzzy controller.Simulations with white noise and road pulse inputs are carried out respectively at different vehicle speeds.The results show that the semi-active suspension system with wheelbase preview-based fuzzy neural network control can effectively suppress vehicle vibration compared to traditional passive suspension and the vertical accelerations at both the mass center of sprung mass and the point on car body vertically corresponding to rear wheel center are significantly reduced compared to fuzzy controlled semi-active suspension,well improving the ride comfort of vehicle.